EABER WORKING PAPER SERIES

Transcription

1 EAST ASIAN BUREAU OF ECONOMIC RESEARCH EABER WORKING PAPER SERIES PAPER NO.33 TRADE AND INVESTMENT PATTERS IN ASIA: REGIONALISATION OR GLOBALISATION? PREMA-CHANDRA ATHUKORALA ARNDT CORDEN DIVISION OF ECONOMICS, RESEARCH SCHOOL OF PACIFIC AND ASIAN STUDIES, AUSTRALIAN NATIONAL UNIVERSITY PAPER PREPARED FOR DISCUSSION AT THE NCAER/EABER CONFERENCE ON THE MICROECONOMIC FOUNDATIONS OF ECONOMIC POLICY PERFORMANCE IN ASIA NEW DELHI 3 4 APRIL 2008 East Asian Bureau of Economic Research. EABER SECRETARIAT CRAWFORD SCHOOL OF ECONOMICS AND GOVERNMENT ANU COLLEGE OF ASIA AND THE PACIFIC THE AUSTRALIAN NATIONAL UNIVERSITY CANBERRA ACT 0200 AUSTRALIA

2 Trade and Investment Patterns in Asia: Regionalisation or Globalisation? Prema-chandra Athukorala Arndt Corden Division of Economics Research School of Pacific and Asian Studies Australian National University Draft paper for presentation to the symposium on Micro-Economic Foundations of Economic Policy Performance in Asia, organized by the National Council for Applied Economic Research and the East Asia Bureau of Economic Research, New Delhi, 3-4 April 2008.

3 1 Trade and Investment Patterns in Asia: Regionalisation or Globalisation? 1 1 Introduction The purpose of this paper is to document and analyse the emerging patterns of international trade and cross-border direct investment in Asia, with a view to informing the contemporary debate on growth dynamism and the process of intra-regional versus inter-regional economic integration. The paper aims to add new insights into the sizeable existing literature on this subject in relation to three major areas. First, a key theme running though this paper is the implications of international fragmentation of production that is, the geographic separation of activities involved in producing a good (or service) across two or more countries for the debate on regional versus global integration of these countries. Second, as an extension to this theme, particular attention will be placed on the debate as to whether the emergence of China as the world s fastest growing industrial economy will crowd out other countries opportunities for integrating into the regional and global economy through fragmentation-based specialization. Third, an attempt is made to examine the emerging complementarities between trade and investment patterns, an aspect which has so far remained virtually overlooked but is vital for understanding the on-going process of economic integration in the region. For the purpose of the study Asia is defined here to encompass the economies of South and East Asia. East Asia includes Japan, and developing East Asia (DEA), which covers the newly industrialized economies (NIEs) in North Asia (South Korea, Taiwan and Hong Kong), China and members of the Association of Southeast Asian Nations (ASEAN). Developing Asia (DA) refers to South and East Asia except Japan. Hong Kong and China are treated as one geographical entity, while reporting data separately for the two economies for comparative purposes. This is justified not only because Hong Kong was reverted back to Chinese sovereignty, but also because the two economies 1 This paper draws on my on-going collaborative research with Hal Hill.

4 2 have increasingly been closely interlinked through trade and investment following China s market oriented reform initiated in the late 1970s. To gain perspectives, the Asian experience is examined in the wider global context. The results suggest that, while fragmentation-based specialisation is now a global phenomenon, it is far more important for economic growth and structural transformation in the East Asian economies than elsewhere. Notwithstanding recent market oriented reforms, India still remains a minor player in this most dynamic area of trade expansion. The degree of dependence of the East Asian economies on this new form of international specialisation is proportionately larger in this region compared to North America and Europe. MNE from America, Europe and Japan, and more recently MNEs from the East Asian NIEs have played a pivotal role in linking the countries in the region to regional and global production networks. A highly important recent development has been the rapid integration of China into regional production networks as a major final assembler based on parts and components imported from the rest of East Asia: an important counterpoint to the popular belief that China s global integration would crowd out other countries opportunities for benefiting from rapidly growing fragmentation trade. These developments do not, however, mean that production fragmentation has contributed to lessening the regions dependence on the global economy. On the contrary, the region s growth dynamism based on vertical specialisation is deeply dependent on its extra regional trade in final goods, and this dependence has, in fact, increased over the years. The on-going process of production fragmentation has thus strengthen the case for a global, rather than a regional, approach to trade and investment policy making in countries in the region. The remainder of the paper is organized as follows. Section 2 examines general patterns of trade since 1970, for Asia as a whole, for the major sub-regions, and by country. This examination includes trade flows over time in aggregate, by major partners, and by major commodity groups, paying particular attention to the phenomenon of network trade based on international fragmentation of production. Central to the discussion in this section is the implications network trade for the relative importance of

5 3 intra-regional versus global economic integration, and the way in which latecomers in the region are hooking into the growth process through global integration. Section 3 investigates investment patterns, focusing on the comparative performance of Asian developing countries attracting and managing foreign direct investment (FDI), and the FDI-trade nexus. The final section summarizes the key findings and draws out some general inferences. 2 Trade Patterns The analysis in this section is based on data compiled from the UN Comtrade database. In order to assess the magnitude and nature of trade within global production networks, it is necessary to separate parts and components (henceforth referred to as components for brevity) from final (assembled) products in reported trade data. We do this through a careful disaggregation of 5-digit level data based on the Revision 3 of the Standard International Trade Classification (SITC, Rev 3) of the United Nation trade data reporting system (See Appendix). In its original form (SITC, Rev. 1), the UN trade data reporting system did not provide for the separation of parts and components from final manufactured goods. The version introduced in the late 1970s (SITC, Rev. 2), which was fully implemented by most countries only in the early 1980s, adopted a more detailed commodity classification that provided for the separation of parts and components within the machinery and transport sector (SITC 7). However, considerable overlap between some advanced-stage assembly activities and related final goods within the sector made it difficult to undertake a precise separation of fragmentation-based trade from total trade (Ng and Yeats 2001). Revision 3, which was introduced in the mid-1980s, marked a significant improvement over Revision 2. In addition to redressing the issue of overlap within SITC 7, it provided for the separation of parts and components trade in the miscellaneous goods sector (SITC 8). These two sectors together accounted for around 70 per cent of total world manufacturing trade (defined as goods belonging to SITC 5 through 8 less SITC 68 (non-ferrous metals)) during the period under study. For the purpose of analyzing overall trade trends and changes in commodity composition we

6 4 combine the data reported under the Revisions 2 and 3 of the Standard International Trade Classification (SITC) for the period from 1969/70 to 2005/6. The separation of components from reported trade data is however possible only from 1992 when almost all countries reporting to the UN trade system had adopted the revised reporting system. The data are tabulated using partner-country records, which are considered more appropriate compared to the corresponding reporter records for analysing trade patterns for a number of reasons (Ng and Yeats, 2003, Appendix 1, Feenstra et al., 2005). Partner-country records are admittedly less susceptible to double counting and erroneous identification of the source/destination country in the presence of entrepot trade compared to data based on reporting country records (e.g., China s trade through Hong Kong and Indonesia s through Singapore). Also, some countries fail to properly report goods shipped from their own export processing zones. These exports are simply lump these exports into one highly aggregated category of special transactions under SITC 9. While no fully satisfactory solutions exist for these problems, it is generally believed that data compiled from importer records are less susceptible to recording errors and reveal the origins and composition of trade more accurately since there normally are important legal penalties for incorrectly specifying this information on customs declarations. Partner country records also permit covering trade of countries which are not covered in the UN data system. Among the countries covered in this study, Taiwan is not covered in the UN data system and Vietnam has just begun to make data available according to the standard UN format. In order to achieve comparability across countries, data on oil and gas trade (SITC 3) are excluded from the coverage of merchandise trade. Among the Asian countries covered in this study, oil and gas account for a significant share of exports only in Malaysia and Indonesia. 2 In both countries there has been a dramatic shift in export 2 The following statistics reveal this declining trend: Share of oil and gas in merchandise exports (%) 1969/ / / /06 Indonesia Malaysia

7 5 composition of exports away from crude oil and non-oil primary products to manufacturing. The date are in current US$ terms. Trends Table 1 summarises data on the relative trade performance of Asian countries in world trade. The combined share of Asian countries in world non-oil exports recorded a threefold increase, from 11.1% to 33.4%, between 1969/70 and 2005/6 (Table 5). 3 The region accounted for over 40% of total increment in world exports over this period. East Asia dominated this impressive export growth story, accounting for over 95% of the total regional trade. Within East Asia, the combined world market share of ASEAN increased persistently from 2.2% in 1969/70 to 6.3% in 2005/6, but these countries still account for less than a fifth of total Asian trade. Notwithstanding notable export expansion in recent years, South Asia still accounts for a mere 1.4% of total world trade, equivalent to less than 5% of Asia s total trade. Among the nine largest DEA economies only Hong Kong, Indonesia and the Philippines have smaller world trade shares than India, by far the dominant South Asian economy. In 1969/70 China s world export share (0.8%) was at par with that of India, but was more than 10 times larger in 2005/6 (India: 1.1%, China 12.3%) In the 1970s and 1980s, Japan dominated the region s trade, accounting for over two-thirds of its exports and imports. The picture has changed dramatically over the past two decades, as DEA s share increased from 13.4% in 1989/90 to 25.1% in 2005/06. The rise of China has been the dominant factor behind this structural shift, but the other countries in the region have also increased their world market shares. Thus, on first inspection, there is no indication of China crowding out its neighbours. We return to this issue below. In the global context, Asia s market share gains have come predominantly at the expense of developed countries. The combined share of other 3 Hereafter, we will use the terms total world exports/trade and total world non-oil exports/trade interchangeably and to mean the same thing. Trade and investment magnitudes throughout the paper are measured in current US dollars unless otherwise indicated. Data are presented as two-year averages to smooth out the impact of yearly fluctuations in trade.

8 6 developing countries (that is, all developing countries less Asian developing countries) has increased persistently, though of course at a slower rate than DEA. Commodity Composition Rapid export growth in Developing Asia (DA), mainly driven by the DEA group, has been underpinned by a pronounced shift in export structure away from primary commodities and toward manufactures (Tables 2 and 3). By 2005/06 manufactures accounted for 92% of total exports from Asia, up from 78.3% three decades ago. Given the nature of their resource endowments, Japan and the four Asian NIEs (Hong Kong, Taiwan, Korea, and Singapore) relied very heavily on manufacturing for export expansion from the very beginning. However, beginning in the 1970s, a notable shift towards manufacturing is observable across all countries, at varying speeds and intensity. Between 1979/80 and 2006/7 the share of manufacturing in total exports of developing Asian countries increased from 63.0% to 92.0%. The shares of ASEAN and South Asia respectively increased from 31.5% to 67.3% and 55.5% to 80.3% between these two time points. Among individual countries Indonesia, Vietnam, and Pakistan (and of course the very small late-comer Indo China economies) have a significantly lower share of manufactures in their exports, reflecting both their comparative advantage and their later adoption of export-oriented industrialization strategies. Asia s share in total world manufacturing exports increased from 19.5% in 1979/80 to 36.6% in 2005/6. This increase came entirely from the DEA economies, since the share of Japan fell sharply over this period (from 11.4% to 7.8%) and South Asia still accounted for a tiny share, around 1%, at the end of the period (Table 6, Figure 1). China s rise has been the key factor behind the rapid growth of manufacturing exports from developing Asia, but exports from Taiwan, Korea, and the ASEAN countries have also recorded impressive growth. Within manufacturing, machinery and transport equipment (SITC 7), in particular the three sub-categories of information and communication technology products 4 therein, 4 Hereafter referred to as ICT, comprising the sum of office machinery (SITC 75), telecommunication and sound recording equipment (SITC 76), and electric machinery (SITC 77).

9 7 have played a pivotal role in this structural shift. The share of Asia in world machinery and transport equipment exports increased from 11.9% in 1989/90 to 41.3% in 2005/06, with DEA accounting for over four-fifths of the increment. By 2005/06, over 67% of total world ICT exports originated from Asia, up from 49% in 1989/90. The share of machinery and transport equipment in the export structures of some of the more industrialized economies of East Asia is particularly high. By contrast, that for Indonesia, Vietnam and all of South Asian is much smaller (Table 2). Asia s share in world exports of the other main product categories has also increased over time, though at a slower rate. Of particular interest here is the notable increase in region s share in miscellaneous manufacturing. This mostly consists of standardized labour-intensive manufactured goods, in particular clothing and footwear. China has accounted for much of this increase but, in contrast to ICT exports, the geographic participation has been broader. A number of low-wage countries in Southeast and South Asia, including Indonesia, Vietnam, India, Sri Lanka, Bangladesh, and Cambodia (the latter included under Other ASEAN countries ) have all recorded impressive gains in market share. Network Trade The fast growth of machinery trade has been driven by rapid growth of international fragmentation of production in world trade and the increasingly deep integration of East Asian countries into the global production networks (Ng and Yeats 2003, Athukorala 2006 and 2008, Kimura 2006). International production fragmentation ( fragmentation, for short) has been an important feature of the international division of labour since about the mid-1960s. The role of East Asia in fragmentation-based global production networks began in 1968 with the arrival in Singapore of two US companies, National Semiconductors and Texas Instruments to assembly semiconductor devices. By the beginning of the 1970s Singapore had the lion s share of offshore assembly activities of the US and European

10 8 semiconductor industries. Virtually every international electronics producer was present in Singapore by the mid-1980s, when the hard disk drive assemblers entered the country further boosting its role as a global assembly centre. During the next five years semiconductor production declined in relative importance, and computer peripherals, especially hard disk drives and computers became the more important part of the islands electronic industry. By the late 1980s, Singapore was the world s largest exporter of hard disk drives, accounting for almost half of world production (McKendrick et al., 2000). As early as 1972 the MNEs with production facilities in Singapore began to relocate some low-end assembly activities in neighbouring countries (particularly in Malaysia, Thailand and the Philippines) in response to rapid growth of wages and land prices. Many newcomer MNEs to the region also set up production bases in these countries bypassing Singapore. By the late 1980s this process had created a new regional division of labour, based on skill differences involved in different stages of the production process and relative wages, and improved communication and transport infrastructure. At the time, there was a widespread concern in policy circles in Singapore that the regional spread of MNE operations in electronics industry could be at the expense of Singapore. However, the subsequent developments vividly demonstrated that the larger the scale and scope of electronic industry [which produces a wide range of heterogenous end-products, each of which needs a large number of equally heterogenous components in its manufacture] in Southeast Asia, the greater the economies of scale and more the opportunities for specialisation for all participating countries (Goh, 1990). More recently, regional production networks have begun to expand to Vietnam (Athukorala, 2007, Chapter 10). 5 Despite obvious advantageous in terms of location and relative wages, Indonesia has so far failed to benefit from this new form of international specialisation because of the unfavourable domestic investment climate (Athukorala, 2006a). 5 Until recently, the fledgling electronics industry in Vietnam was largely dominated by small companies from newly industrialized countries in East Asia, with the sole exception of Fujitsu which operated a medium-size assembly plat in Ho Chi Ming City. On 28 February 2006, Intel Corporation, the world s largest semiconductor producer, announced that it will invest $300 million to build a semiconductor testing and assembly plant (with an initial workforce of 1200) in Ho Chi Ming City as part of its worldwide expansion of production capacity.

11 9 The continued attraction of East Asia as a location of assembly activities seems to have been underpinned by a number of factors. First, despite rapid growth, manufacturing wages in all ASEAN countries except Singapore still remain lower than or comparable to those in countries in the European periphery and Mexico. 6 Moreover, significant differences in wages among the countries within the East Asia region have provided the basis for rapid expansion of intra-regional product sharing systems, giving rise to increased cross-border trade in parts and components. Second, the relative factor cost advantage has been supplemented by relatively more favourable trade and investment policy regimes, and better ports and communication systems that facilitate trade by reducing the cost of maintaining services links. Third, as first-comers in this area of international specialisation, ASEAN countries (in particular Malaysia, Singapore and Thailand) seem to offer considerable agglomeration advantages for companies that are already located there. Site selection decisions of MNEs operating in assembly activities are strongly influenced by the presence of other key market players in a given country or neighbouring countries. Against the backdrop of a long period of successful operation in the region, many MNEs (particularly US-based MNEs) have significantly upgraded technical activities of their regional production networks in ASEAN and assigned global production responsibilities to affiliates located in Singapore and more recently also to those located in Malaysia and Thailand (Borrus et al., 2000; McKendrick et al., 2000). All in all, the ASEAN experience seems to support the view that MNE affiliates have a tendency to become increasingly embedded in host countries the longer they are present there and the more conducive the overall investment climate of the host country becomes over time (Rangan and Lawrence, 1999; Athukorala and Yamashita, 2006). 6 Average annual compensation (Salary/wage plus other remuneration) per worker (US$) in selected countries: China 1835 (2001), Indonesia 880 (2000), Philippines 2965 (2000), Thailand 3345 (1994), Malaysia 4380 (2000), Vietnam 650 (2000), Taiwan (1997), Korea (2000), Singapore (2000), Poland 2502 (2000), Hungary 2898 (2000), Czech Republic 4150 (1998), Mexico 8050 (2000) (Source: China: China Statistical Press (2003) (average wage for Beijing, Tianjin, Shanhai, Zhejiang, Liaoning and Guangdon); Vietnam, General Statistical Office 2000; other countries: Nicita and Olarreage (2006), Statistical Appendix).

12 10 Over the years Singapore s role in regional production networks has gradually shifted from low-skill component assembly and testing to component design and fabrication and providing head-quarter services for production units located in the neighbouring countries. Singapore s attractiveness as the regional centre of cross-border production networks has been continuously enhanced by the policy emphasis of the government on infrastructure development, expanding the human capital base, maintaining labour relations in a manner highly conducive for international production, and sound macroeconomic management (McKendrick et al., 2000; Brown and Linden, 2005). While MNEs from the US dominated the scene at the formative stage of the spread of assembly activities in the late 1960s, Japanese and Western European MNEs have become increasingly involved since the late 1970s. More recently, MNEs from more advanced developing countries, notably those from the East Asian NIEs have also joined this internationalization process. In response to rising domestic wages, the growing reluctance of domestic labour to engage in low paid blue-collar employment, and stringent restrictions on the importation of labour, firms in electronics and other durable consumer goods industries in the East Asian NIEs have begun to produce components and subassemblies in low-cost neighboring countries. There is no unique way to measure the intensity of fragmentation-based specialization. However, for the purpose of time-profile analysis of the phenomenon, the best available indicator of is the share of parts and components in total recorded trade in machinery and transport equipment (Yeats 2000, Athukorala 2006a, Jones, Kierzkowski Lurong 2005). 7 Data on world market shares of machinery exports disaggregated into components and finals goods (reported total exports parts and components) and the relative importance of components in total machinery exports are summarised in Table 4. 7 Henceforth, for the sake of brevity, we use the term components in place of parts and components and machinery in place of machinery and transport equipment.

13 11 The share of East Asia in world component trade increased from 34% in 1989/90 to 40% in 2005/6. This resulted from an increase in the share of DEA from 16.5% to 28.1%, which more than counterbalanced a sharp decline in the share of the one-time regional giant, Japan, from 17.8% to 11.3%. Within DEA, all countries covered in our data tabulations have recorded increases in world market shares, with the ASEAN countries exhibiting faster increases compared to the regional average. India remains a tiny participant in the global production networks even though it has great potential to benefit from this new form of international specialization, given the relatively low-cost and trainable labour, and its location in a region that has become the growth centre of component production and assembly in the world. In 2005/6 India accounted for a mere 0.3% of component exports and 0.2% of final good exports in world machinery trade. The dependence of the East Asian countries as a group on component trade is much higher compared to all other regions in the world. In 2005/6, components accounted for 34.7% of total machinery exports of these countries. Within East Asia, the ASEAN countries, in particular Malaysia, Philippine, Singapore and Thailand, stand out for their heavy dependence on product fragmentation for export dynamism. In 2005/6, components accounted for 58.4% of the total exports of the six ASEAN countries, up from 46.7% in 1989/90. For instance the combined export share of the six main member countries of the ASEAN Free Trade Area (AFTA) more than doubled (from 5.9% to 10.4%) between these two years. In the 1990s there was growing concerns in policy circles in East Asia that the formation of NAFTA and the integration of some new countries emerging from the former Soviet Union with the rest of Europe would adversely affect the relative position of developing East Asian countries in world assembly activities. Indeed, in addition to the new tariff concessions, proximity to industrial countries and relatively low wages by regional standard (though not compared to some of the East Asian countries) can be considered as added advantages of these countries compared to East Asian countries in production fragmentation based international specialisation (Egger and Egger 2005, Ng

14 12 and Yeats 2003, Kierzkowski 2001). However, this prediction does not seem to have materialized world market shares of Mexico and rest of Europe (EUT less EU) have increased, but at a much slower rate than that of developing East Asia. It seems that in spite of geographical proximity and tariff concessions under FTAs, US producers still find East Asia as a more attractive location for fragmentation-based international exchange for the reasons discussed at the outset in this section. The growing importance of China is component trade particularly noteworthy. Between 1992/3 and 2005/6, the share of China in total world exports and imports of component increased from 1.4% to 7.3%, and in total imports from 2.5% to 9.8% respectively. A comparison of the data on the share of components in total exports and imports (the last 2 columns in Table 4) points to an important difference between China and the other Developing East Asian Countries (DEACs). In China components accounts for a much larger share of imports (60.4% in 2005/6) compared to that in exports (34.8%). In other DEAs the percentage shares are broadly similar on the import and export side. This difference seems rapid expansion of China s role in world trade has brought about a notable shift in the regional division of labour, with the other DEACs, in particular ASEAN countries, playing an increasing role in producing parts and components for the rapidly growing final assembly activities in China. 8 China in regional production networks The data reported in Table 6 illustrates the growing importance of East Asia in China s machinery trade, focusing separately on trade in components and final products. The data clearly reflect the evolving role of China as an assembly centre within the East Asian region. The share of East Asia in total parts and component imports to China has increased sharply. By 2004/5, over two thirds of total components imports to China originated in the region. By contrast, only 47.8% of Chinese final products found markets in these countries. The share of Japan in component imports to China declined from 26.8% in 1992/3 to 22.6% 8 The simple point here is that specialisation in component production/ assembly involves both importing and exporting of components within global production chains, involving multiple border crossing of components as countries add value at different slices of the value chain. By contrast final assembly in China involves importing components and exporting the assembled final products largely to North America, the EU and other third country markets..

15 13 in 2004/5, but all other East Asian countries have gained market shares. The combined share of imports from Southeast Asian countries recorded the sharpest increase, from a mere 1 % in 1992/3 to 16.0% in 2004/5. Among these countries, import shares of Malaysia and the Philippines have increased at a faster rate compared to that of Singapore. The increase in market of the countries in the region has been reflect in a sharp decline in China s extra-regional imports of components. For instance, import shares of the US and EU declined from 10.5% to 7.0%, and from 19.4% to 10.7% between these two time points. This new China dimension of regional production networks in East Asia is further illustrated by the data reported in Table 5. Compared to the increased intraregional dependence in component imports of China, geographic composition of its final goods exports is characterised by a clear extra regional bias. In 2004/6, 62% of total final product exports found market outside the East Asian region, up from 45% in 1992/3. The share of final exports to East Asia and DEA declined from 55.0% to 38.0%, and 49.5% to 26.5% between 1992/3 and 2004/6. Final goods exports to ASEAN accounted for around 5% throughout this period, whereas the share of component imports from ASEAN increased from less than 2% in 1992/3 to over 16.0 in 2005/06. These differences in the geographic patterns of imports and exports reflects the increasingly important role played by China as a final product assembler for advanced-country markets using parts and components procured from countries in the region, with Southeast Asian countries among them playing an increasingly important role. The data also point to an increase in twoway trade in components between China and the rest of East Asia. For instance the share of Southeast Asia in total component exports of China increased from 5.8% to 13.0% accompanied by to 16% an increase in the region s share of Chinese exports from 1% between 19992/3 and 2004/6. This pattern reflects the occurrence of multiple border crossing of components at different stages in the production process, a well documented feature of international production fragmentation (Brown and Linden 2006). However, reflecting China s preeminent role as the global assembly centre in the region, the total value of component exports to the other East Asian countries (and also to the entire world) is much smaller compared to imports from these countries (in 2005/6: imports

16 14 from and export to the rest of East Asia: US$ 125 billion and US$ 74). In 2005/6, China ran an overall trade deficit (as a percentage of exports) of -62% in components trade and a surplus of 103% in final good trade in machinery. Almost 90% of the deficit of component trade was with the rest of East Asia whereas the share accounted by intraregional exports in total surplus in final trade was around 25%. Overall patterns of fragmentation trade Table 6 compares geographic patterns of total manufacturing trade and trade in components. The data vividly demonstrate the East Asia s peculiar heavy reliance on fragmentation-based international exchange in the global context. For instance, in 2005/6 intra-regional exports accounted for 44.4% of total manufacturing exports from East Asia. The comparable figure for component exports was 60.9%. Intra-regional regional share in component imports is even larger. These patterns are in sharp contrast to those in NAFTA and EU15 (as well as in overall global trade). Moreover the intra-regional shares in total component imports and exports have grown faster between 1992/3 and 2005/6 compared those in total imports and exports. The increase in component intensity has been particularly in ASEAN s trade with the other developing East Asian countries, China in particular. Korea and Taiwan are also involved in sizable cross border trade with the other countries in the region. For all East Asian countries, the share of components in both intra-regional exports and imports have increased at a much faster rate than in exports to and imports from countries outside the region. These patterns are in sharp contrast to those observed for NAFTA and EU15 (as well as total global trade). In both these geographical entities, the shares of intra-regional trade in total manufacturing trade (both on export and import sides) and component trade imports are broadly similar in magnitudes. Direction of Trade: Regionalisation or globalisation? There is a vast literature on what may be termed standard trade data analysis, that is, essentially based on the traditional notion of horizontal specialisation, in which trade is an exchange of goods that are produced from start to finish in just one country. This literature unequivocally points to a persistent increase in intra-regional trade in East Asia,

17 15 whether or not Japan is included or not, from about the early 1980s. 9 This evidence figures prominently in the current regional debate concerning the establishment of regional trading arrangements covering some or all countries in East Asia. In particular, the proponents of the proposal to expand AFTA to encompass Japan, China and South Korea (the ASEAN+3 proposal), and more broadly towards an Asian Economic Community, often refer to deepening economic interdependence, as reflected in intraregional trade among these countries, as evidence of its likely success. Increasing trade integration is also cited as an indicator of the potential benefits of monetary integration in the region (Kwan 2001). However, the discussion so far in this section on the emerging patterns of intraregional component trade casts doubts on the validity of these inferences. We have noted two important peculiarities of trade patterns in East Asia compared to global trade and trade patterns for the EU and NAFTA countries. First, component trade has played a much more important role in trade expansion in East Asia relative to the overall global experience and that of countries in other major regions. Second, trade in components accounts for a much larger share in intra-regional trade than it does in trade with the rest of the world. Given these two peculiarities, trade flow analysis based on reported trade data is bound to yield a misleading picture as to the relative importance of intra-regional trade relations (as against global trade) for growth in East Asia, and also ASEAN. This is because growth based on assembly activities depends on the demand for final goods, which in turn depends increasingly on extra-regional growth. Data reported in Table 7 on intra-regional shares of trade in total manufacturing, components and final goods for various regional economic groupings help understand this important point. Intra-regional shares computed using unadjusted data for total trade (exports + imports) (the commonly used indicator of regional economic integration in East Asia) reported in Table 7 affirms the received view that Asia, in Particular East Asia, has 9 See for example Kwan 2001; Drysdale and Garnaut 1997; Fujita 2008, Petri 1993, Lee and Roland-Holst 1989.

18 16 become increasingly integrated trough merchandise trade. The share of intra-regional trade in East Asia increased from 34.2% in 1979/80 to 47.3% in 1992/3 and then to 54.0% in 2005/6. For Developing East Asia, the magnitudes of the comparable fifires are a little lower, but the degree of increment over the time has been faster; from 21.2% in 1979/80 to 42.6% in 2005/6. By 2005/6, the degree of trade integration in East Asia had surpassed that of NAFTA (40.7%) and was rapidly approaching that of EU15 (59.0%). When we shift our focus from total trade to export and import trade, we notice a startling asymmetry in the degree of measured trade integration in East Asia (and all the sub-regions there is), which has surprisingly been ignored in the previous literature. Unlike in EU and NAFTA, in East Asia the increase over time in intra-regional trade ratio has emanated largely from rapid increase in intra-regional imports; intrea-regional exports expansion has lagged behind persistently. In 2005/06 intra-regional import flows amounted to 68.9% of total manufacturing imports of East Asia, up from 38.9% in 1979/80. Intra-regional share in total regional exports was, however, significantly lower, 30.6% in 1979/80 and 45.4% in 2005/06. In other words, the region is much more heavily dependent on extra-regional trade for its growth dynamism than is (misleadingly) suggested by the total regional trade share, and this dependence has remained virtually unchanged for the last decade. In other world, the aggregate intra-regional trade share used widely in policy forums camouflage the continuing importance of extra-regional trade fo growth dynamism in East Asia. Until about the mid 1990a, this asymmetry in intra-regional trade was largely a reflection of the unique nature of Japan s involvement in fragmentation trade in East Asia. Japan s trade relations with the rest of East Asia is predominantly in the form of using the region as an assembly base for meeting demand in the region and, more importantly for exporting to the rest of the world. Japan has persistently maintained a trade surplus with all East Asian countries in both total manufacturing trade and trade in component, of which the latter is much larger (data not reported for brevity). Over the past decade or so, China s emergence as the final goods assembly center in the region has further contributed to this asymmetry. As already noted, rapid expansion in export-

19 17 oriented assembly in China relies heavily on intra-regional imports of components where as final goods imports from China are largely destined to extra regional markets. More importantly, the picture revealed by the total (export + imports) intraregional trade ration change dramatically when we net out parts and components from published trade data and focus instead on final trade. The two alternative estimates (original and adjusted) are vastly different for East Asia, particularly for DEA and ASEAN. Both the level of trade in the two given years and the change over time in intraregional trade shares are significantly lower for estimates based on final trade. The intraregional share of final manufacturing trade in East Asia increased only marginally between 1992/3 and 2005/6 (from 46.2% to 47.5%) in sharp contrast to a notable increase (from 47.3% to 54%) recorded by the conventionally used trade share. While the difference between intra-regional shares of final and total trade is observable for both exports and imports, the magnitude of the difference is much larger on the export side. The difference in magnitude between regional trade shares estimated in gross and net terms is much larger for developing East Asia and ASEAN compared to estimates for the entire region. In 2004/06 only 29.7% of final goods exports from developing Asia found markets within the region, compared to 39.1% in total exports. For AFTA the relevant figures were 14% and 19.7%, respectively. It is also interesting to note that, unlike in the case of East Asia (or developing East Asia and AFTA), the estimated intra-regional trade share for NAFTA, the EU and the other regional groupings are remarkably resilient to the inclusion or exclusion of component trade. In sum, the estimates presented in this section support the hypothesis that, in a context where fragmentation based trade in expanding rapidly, the standard trade flows analysis can lead to misleading inferences regarding the on-going process of economic integration through trade. Production fragmentation leads to double-counting of trade flows in published trade data because goods in process cross multiple international borders in the course of their production sequence. The total amount of trade involving the goods while in process can be a multiple of the final value of that good. Moreover, trade shares calculated using reported data can lead to wrong inferences as to the relative importance of the region and the rest of the world for growth dynamism of a given

20 18 country/region, even controlling for double counting in trade. When data on component trade are excluded from trade flows, our estimates suggest that extra-regional trade is much more important than intra-regional trade for continued growth dynamism of East Asia, both including and excluding Japan. This is because the rate of expansion of component trade depends crucially on the demand for the related final goods. 3 Investment patterns The available data on FDI and MNE operations are generally of poor quality compared to the trade data. The only available time series data relates to only aggregate FDI inflows and out flows. Even these data (readily available from the World Investment Report database of the UNCTAD) suffers from a number of limitations. 10 Not all countries record every component of FDI flows. 11 For most countries, the data series on FDI capture only equity capital and inter-company debt; in fact, the majority of countries do not report data on the third component. There is evidence that the component retained earnings in FDI is positively related to the years of operation of firms in a given country, and that US MNEs have a general tendency to rely more on retained earnings for investment expansion compared to MNEs from other countries (Lipsey 2000). Thus, this problem of data coverage can lead to a considerable underestimation of the actual magnitude of FDI in a given host country, depending on the history of MNE involvement and the source country profile of FDI. Even for the components for which data are available, the quality of data varies considerably across countries. For instance, some countries (such as China and Hong Kong) do not make an adequate distinction between portfolio investment and foreign direct investment. For these reasons, a comparison of data among countries, and even over time for a given country, should be made with 10 For details on the nature and limitations of the World Investment Report FDI data see UNCTAD 2005, Box According to the standard definition, FDI consists of three components. These are: (a) equity capital, that is, the shares owned by the foreign direct investor (MNE) in its affiliates firms; (b) retained earnings, that is, the MNE s share (in proportion to its direct equity participation) of earnings not distributed as dividends by affiliates, or earnings not remitted to the parent company (such retained profits are reinvested by affiliates); and (c) intra-company loans or intra-company debt transactions (except that for working capital) referred to as short- or long-term borrowing and lending of funds between the parent company and affiliated enterprises.

21 19 caution. Moreover, the data coverage tends to vary over time in a given country because of changes made to the data recording system. 12 Data of FDI inflows from the World Investment Report database are summarised in Tables 8 and 9. Total FDI flows to developing Asia increased sharply from an average annual level of $1.5 billion during to $200 billion in The share of Asia in total FDI flows to developing countries almost doubled (from 32.6% to 52.6%) between these two time points. As a share of total global flows, the increase was from 6.7% to 15.3% (Table 12). FDI inflows as a share of GDFCF have been significantly higher than the comparable figure for all developing countries throughout the period , followed by a minor reversal in the pattern during the years of the Asian financial crisis, The average FDI/GDFCF ratio for developing Asia for the entire period was 9.2%, compared to 6.2% for all developing countries and a global average of 7.0%. A notable feature within developing Asia is the dramatic increase in inflows to China. During the first four years of reforms ( ) inflows were rather modest, amounting to about $2 billion. FDI inflows began to gather momentum in the latter half of the 1980s, but were interrupted by the Tiananmen Square incident in Then, from 1991, FDI began to increase dramatically (Wang, 2005). Over the past two decades China has been by far the largest developing country recipient of inward FDI. China s share in total FDI inflows to all developing and transition economies increased from 11% during to 32% during For the six years , China has been the second largest recipient of foreign investment in the world, at about $50 billion per annum and accounting for 7% of total gross inflows, after the USA, which has received about $140 billion per annum, or 13% of total inflows (UNCTAD 2005). China s share in inflows to Asian developing countries increased from 11.4% during to 48.5% during For instance, the Reserve Bank of India broadened the coverage of its FDI estimation procedure in 2003 (with effect from 2000/1 fiscal year) to include retained earnings. According to the revised data for 2000/1 and 2001/2, on average the new component accounted for about 40% of the total reported FDI figures (Reserve Bank of India).

22 20 06, and it has accounted for well over half of the total increment in FDI inflows to the region during this period. Total FDI flows to the ASEAN countries increased sharply from an average annual level of $3 billion in the second half of the 1980s to nearly $30 billion during the six years before the onset of the recent financial crisis. At the time the region accounted for one-fourth of total inflows to developing countries and a little over a third of total inflows to developing Asia. Singapore, Malaysia, Thailand and Indonesia were among the eight largest developing country recipients of FDI. Flows to the Philippines continued to remain low by international standards, but they increased from $40 million to over $1 billion by the mid-1990s. The lion s share of ASEAN s FDI inflows was absorbed by Singapore 50% of the regional total in the 1980s and 40% in the 1990s while Indonesia, Malaysia and Thailand absorbed most of the remainder. In the first half of 1990s, net capital inflows relative to gross domestic fixed capital (GDFCF) stood at over 30% in Singapore, 19% in Malaysia, 10% in the Philippines, and 4% in Thailand. As they opened up, the transition economies of Cambodia, Laos and especially Vietnam recorded healthy FDI inflows from the first half of the 19990s, before faltering during the latter part of the decade. In these countries, FDI accounted for much larger shares of GDFC, reflecting continuing low levels of domestic private investment until recently as a result of the insecure investment climate for local SMEs. 13 The data point to a notable break in investment flows to ASEAN countries following the onset of the financial crisis. Total FDI inflows to the region have declined persistently from about $35 billion per annum prior to 1997 to an annual average of about $24 billion during However, the post-crisis experiences of individual countries vary substantially. Indonesia experience negative FDI inflows until 2004, contributing significantly to the decline in total flows to the region. When the three atypical boom years prior to the onset of the crisis are excluded, owing to the abnormal investor euphoria, there is no discernible break in the trend of FDI inflows to Singapore, Thailand and the Philippines. Flows to the Philippines, the country least affected by the 13 Namely, about 30% in Vietnam, 20% in Laos and 24% in Cambodia.

23 21 crisis among this group, in fact continued to increase rapidly throughout. Net FDI flows to Malaysia declined from $7.2 billion in 1996 to $6.0 billion in 1997, a 24% contraction, and have remained virtually flat at that level from about mid This is contrast to a significant increase in flows to Korea and Thailand. It could well be that the prolonged period of policy and political uncertainty following the onset of the crisis, and widespread market skepticism about the fate of Malaysia s unorthodox reform package introduced in September 1998, may have played a role. The two extreme cases of Indonesia (continuous contraction until 2003) and the Philippine (continuous increase until its own political woes in recent years) clearly suggest the post-crisis decline in FDI inflows to the region was a temporary aberration associated with economic disruption and political turbulence caused by the crisis. Moreover, there is also evidence that the decline in FDI after the onset of the crisis was by and large limited to domestic market-oriented investment, while FDI in export-oriented industries continued to increase throughout the period, boosted by the now highly competitive exchange rates (Athukorala 2003). It is also important to note that the continuation of the crisis-driven decline in FDI inflows to these countries well beyond the period of recovery after the crisis (that is, beyond 2000) was largely a reflection of a large overall decline in global FDI flows during (UNCTAD 2005), and a global downturn in electronics. Total global FDI inflows declined from $134 billion in 2000 to $83 billion in 2001, $72 billion in 2002, and $63 billion 2003, before recovering marginally to $65 billion Total inflows during the four years from 2001 to 2004 were 24% lower than the comparable figure for the preceding four years, Interestingly, FDI inflows to the crisisaffected Asian countries (and to developing Asia in general) seemed to have been remarkably resilient in the face of this massive global contraction. The 1990s saw a marked increase in FDI to India, a trend that represents a clear break from the preceding two decades. India s share of FDI in total developing country inflows increased from 0.4% in the 1980s to over 1.5% in the first two years of the new 14 There has yet to be a comprehensive explanation for this massive contraction in FDI flows, an unprecedented occurrence since 1970, when the Word Investment Report FDI series commenced.